READ ME

This replication archive contains one data set in Stata format: worms_replication.dta

All analysis was done in Stata 15.1

All replication code is written as Stata .do files that call this data set. To start any replication, change the directory in Stata to be the folder that the archive is in. 


SUMMARY STATISTICS: Summary statistics reported on pages 15-20 can be replicated using summary_statistics.do

FIGURE 2: The histogram in Figure 2 is replicated using figure_2_histogram.do

FIGURE 4: There are three cells in Figure 4 created using the three files figure_4_trade.do, figure_4_calories.do, and figure_4_inhabitants10.do

TABLES 1, C.2, and C.3: The results in these tables are created using five different do files (one for each wealth measure). These are:
	results_trade.do (row 1 of each table)
	results_calories.do (row 2 of each table)
	results_city5.do (row 3 of each table)
	results_city10.do (row 4 of each table)
	results_city15.do (row 5 of each table)
These do files run 15 regressions each. The first 5 correspond to the 5 columns of Table 1, the next 5 correspond to the 5 columns of Table C.2, and the final 5 correspond to the 5 columns of Table C.3. The table reports the coefficient and standard error of the interaction term from the regressions.


FIGURE 5: There are five cells in Figure 5 created using the five files Figure_5_trade.do, Figure_5_caloric_potential.do, Figure_5_inhabitants5.do, Figure_5_inhabitants10.do, and Figure_5_inhabitants15.do

TABLE 2: This table reports the results of the three regressions run in table_2_growth.do 

TABLE C.1: This table reports the results from the regressions run in missing.do



	 

